The distribution of guyots and atolls and large volcanic islands on the Pacific plate can be used to outline the likely connection between stresses acting on the plate and the gradual development of large, linear volcanic chains over the past 170 Ma. We construe three general periods with different stress regimes in the history of the Pacific plate. 1) During the Jurassic and Early Cretaceous, the Pacific plate was surrounded by ridge segments and there were no major stress alignments within it. Within-plate volcanism thus assumed the scattered arrangement for the condition of no tectonic stress (1), and the large Magellan and Wake seamount clusters formed. Near the eastern boundaries of the plate, complex and shifting patterns of ridge reorganization dictated formation of very long, splayed, near-axis ridges such as Horizon Guyot and Necker Ridge. 2) At about 90 Ma, the growing middle-aged Pacific plate achieved its first persistent stress regime with the formation of subduction boundaries along its western or northwestern margin. The plate was no longer static but began to move over the asthenosphere and into the mantle. Subduction boundaries and the overall direction of subduction are uncertain, but this imparted a general yet not fully stable component of tension across the plate, producing the NNW Gilbert-Marshall, Line and Emperor Seamount ridges, generally orthogonal to the overall direction of least principal stress. The Line Island seamount chain, being near ridge axes, sustained a variable stress regime. It thus has no age progression of rocks dated between 70-90 Ma (2), great width, and a dual orientations of ridges. 3) By 47 Ma, nearly half of the boundaries of the Pacific plate now were trenches spanning from the Aleutians to New Zealand. In addition, northward migration of the Indian plate and Australia caught a major portion of the westerly moving Pacific plate between the northeast corner of the Tonga Trench and the Aleutians. The plate could not shift laterally in response to whatever was occurring along its eastern spreading boundaries. A very consistent and strong stress regime therefore developed across the Pacific plate with a NNE direction of least principal stress. The change in stress orientation may have taken up to 10 million years, during an interval marked by little or no volcanic productivity at the western end of the Hawaiian chain. Since that time, the predominant alignment of both linear island chains and Puka Puka-type ridges, from the Kodiak-Bowie chain in the Gulf of Alaska to the Louisville Ridge south of the Antarctic convergence, has been orthogonal to this direction. Development of large-volume persistent chains and shorter small-volume chains indicates patterns of differential stress in the plate, variable fertility and geochemistry of the asthenosphere and/or shallow convective overturn of the asthenosphere rather than the action of mantle plumes of different sizes and depths of origin. Tapping of enriched mantle by widespread volcano clusters during the Mesozoic suggests the presence of a shallow asthenospheric source layer rather than multiple narrow conduits.

(1) Hieronymus, C.F., and Bercovici, D. 2000. Earth Planet. Sci. Lett. 181, 539-554.
(2) Davis, A.S., Gray, L.B., Clague, D.A., and Hein, J.R., 2002 Geochem. Geophys. Geosyst. 3: 10.1029/2001GC0000190, 1-28.